1. Field of the Invention
The present disclosure relates to an auto-focus system, more particularly to an auto-focus system using multiple lenses to improve speed and precision of the focusing.
2. Description of the Related Art
Generally speaking, the auto-focus function is one of importance functions of the current camera, such as digital still camera or digital camcorder. By the auto-focus function, user can get the focus length of the lens set of the camera quickly, so that the success rate of shooting picture and quality of the picture can be improved. In addition, the auto-focus function can also correctly track the quickly moving object, to lower the threshold of photograph.
In prior art, the basic operation of the auto-focus function is to automatically control the movement of at least one optical lens elements of the lens of the camera system, whereby the object can be clearly imaged on the image sensor unit. To capture the images of the objects locate at different positions or different depths, the at least one optical lens elements must be moved to different focus positions repeatedly. Therefore, the focus speed of the auto-focus function is an importance factor affecting the user experience. Many companies often highlight the function of quick focus to promote their products.
Common auto-focusing approaches are classified into a two dimensional focusing and a three dimensional focusing. In operation of the two dimensional focusing, the at least one optical lens element of the lens is moved for many times and then the focus status of the lens is determined according to the clear degree of the image received in the image sensor unit after each movement of the at least one optical lens element. Generally speaking, when the lens does not converge well, the image formed on the image sensor unit is blurring and the contrast value of the image is lower. Therefore, a focus area can be assigned in advance, and a contrast value in the focus area of the image received in the image sensor unit is calculated by image process after each movement of the at least one optical lens element. The contrast value can indicate the clear degree of the image. After multiple contrast values are obtained, a contrast value curve can be estimated according to these contrast values, and the position corresponding to the maximum value in the contrast value curve can be defined as the focus position, and the at least one optical lens element is moved to the focus position for completing the focusing.
Other focusing approach is a three dimensional focusing. When the camera has two sets of the lens and the image sensor unit, the camera can generate two images at the same time, and a depth of the object within the two images can be calculated according to the two images. The depth indicates a distance between the camera and the object. At least one of the two lenses is moved to a position corresponding to the depth, so the focusing is completed.
The two dimensional focusing requires moving the optical lens element for many times to obtain multiple contrast values, which results in longer focus time. Compared with the two dimensional focusing, the three dimensional focusing just calculates the depth of the object one time to complete the focusing, and the focus time of the three dimensional focusing is about one tenth of the focus time of the two dimensional focusing, so the three dimensional focusing has advantage of high focus speed.
However, the three dimensional focusing has a drawback that the depth calculation fails easily if a repeated pattern appears in the focus area of the image, and the wrong depth usually causes the failure of focusing.
Therefore, what is need is an auto-focus system which has the advantage of the high focus speed in the three dimensional focusing and also can prevent from the failure of focusing when the repeated pattern appears.